Table 6. Tolerance of HbT images recovered without using structural priors to various experimental errors in all phantom cases.
| (a) Metric 1: Overall RMSE ≤ 2 µM | |||||||
|---|---|---|---|---|---|---|---|
| Case 1 | Case 2 | Case 3 | Case 4 | Case 5 | Mean | std | |
| Offset in x-direction (mm) | 0.25 | 0.01 | 0.73 | 0.71 | 0.98 | 0.54 | 0.39 |
| Source plate rotation (°) | 0.11 | 0.01 | 0.31 | 0.27 | 0.40 | 0.22 | 0.16 |
| Source plate L-R tilt (–°) | −0.49 | −0.02 | −1.16 | −0.97 | −1.30 | -0.79 | 0.53 |
| Source plate L-R tilt ( + °) | 0.56 | 0.01 | 1.02 | 1.02 | 1.55 | 0.83 | 0.58 |
| Source plate N-C tilt (°) | 0.11 | 0.00 | 0.29 | 0.28 | 0.60 | 0.26 | 0.23 |
| Breast expansion (%) | 0.50 | 0.02 | 1.30 | 1.13 | 1.66 | 0.92 | 0.66 |
| Thickness increase (%) | 1.09 | 0.04 | 1.28 | 1.21 | 1.87 | 1.10 | 0.66 |
| Cross-talk (‰) | 0.10 | 0.00 | 0.13 | 0.13 | 0.24 | 0.12 | 0.09 |
| Light piping (‰) | 0.41 | 0.11 | 0.80 | 0.84 | 1.71 | 0.77 | 0.60 |
| (b) Metric 2: R, HbTt, or HbTn change ≤ 5% | |||||||
| Case 1 | Case 2 | Case 3 | Case 4 | Case 5 | Mean | std | |
| Offset in x-direction (mm) | 0.77 | 1.81 | 1.14 | 1.79 | 1.28 | 1.36 | 0.44 |
| Source plate rotation (°) | 0.38 | 0.79 | 1.47 | 1.72 | 1.38 | 1.15 | 0.55 |
| Source plate L-R tilt (–°) | −0.80 | −0.60 | −0.57 | −0.64 | −0.61 | -0.64 | 0.09 |
| Source plate L-R tilt ( + °) | 0.88 | 0.63 | 0.59 | 0.68 | 0.63 | 0.68 | 0.12 |
| Source plate N-C tilt (°) | 0.51 | 0.41 | 0.40 | 0.42 | 0.46 | 0.44 | 0.05 |
| Breast expansion (%) | 4.97 | 4.19 | >10.00* | >10.00* | >10.00* | 4.58** | 0.55** |
| Thickness increase (%) | 3.36 | 1.76 | 1.49 | 1.39 | 1.41 | 1.88 | 0.84 |
| Cross-talk (‰) | 0.32 | 0.72 | 0.29 | 0.19 | 0.44 | 0.39 | 0.20 |
| Light piping (‰) | 0.96 | 1.94 | 1.16 | 0.85 | 1.77 | 1.34 | 0.49 |
*:
Tolerance values are beyond the maximal tested experimental error levels.
**:
Statistics are determined by excluding tolerance values marked by *.